Deliverable-Type: lead-scoring-model
Title: Dual-Axis Lead Scoring Model (Marketo/SiriusDecisions pattern)

## Explicit-Fit Criteria (firmographic)

| Criterion | Points |
|---|---|
| Job title contains "founder"/"owner" | +20 |
| Company revenue > $1M (enriched) | +15 |
| Industry matches ICP list | +10 |
| Personal email domain (gmail/yahoo) | -10 |
| Company size < 2 employees, no site | -15 |

## Implicit-Behavior Criteria (engagement)

| Behavior | Points |
|---|---|
| Opened last 3 emails | +5 each (max +15) |
| Clicked a pricing/offer link | +20 |
| Attended a live call | +25 |
| Unsubscribed from list | -25 |
| No engagement in 30 days | decay: -5 per week of silence, floor at 0 |

DECAY rule: all implicit/behavior points decay at -5 per rolling 7-day window of no activity, floor at 0. Explicit/fit points do not decay (firmographic facts don't expire on the same clock).

MQL threshold: total score >= 50 -> routes to outreach-operator queue with a 60-minute SLA for first human touch.

Review cadence: score weights and the MQL threshold are reviewed monthly against closed-won conversion data.

Goodhart note: the gameable path is reps/automations "opening emails" via auto-preview bots to inflate implicit score without real intent. Counter-metric: pair implicit-score volume against MQL->SQO conversion rate — if score volume rises while conversion falls, the score is being gamed and weights get re-audited.